Sara Valentina Talat on AI Orchestration for Business

September 17, 2026
Sara Valentina Talat · BRIICK
Co-Founder & CEO · BRIICK
AI InterConnect Podcast · Published September 17, 2026

Sara Valentina Talat is the co-founder and CEO of BRIICK, which builds AI agents and an AI orchestrator to help service businesses automate customer interactions and operational workflows across their existing systems. She tells host Laurent Cochet how a career in customer engagement led her to the company, how she picks the first workflows to automate, and what a CEO should ask before letting AI talk to customers.

From customer engagement to an AI orchestrator

Talat spent 16 years working on customer journey automation, and she says everything starts with the customer. A business succeeds when every customer feels special and well looked after, which only happens when it connects the dots across the full experience. The opposite, in her view, is ads and discount vouchers for a product someone already bought at full price, or a sales call while a customer service ticket is still open. Personalization, she says, is the secret of customer engagement, and generative AI can be a superpower for customer experience or a nightmare.

She separates adopting a tool, which is just something else to learn, from becoming an AI-native company, where every piece of business knowledge and every customer conversation sits in a database AI can read. Humans and AI employees then work from the same knowledge. Instead of talking to three colleagues and opening several applications to understand one customer, a person asks a machine and gets the full picture in seconds.

Each AI agent can do only one job, she says; a car cannot cook dinner. So a business needs separate agents for customer service, lead qualification or document creation. Nobody wants to talk to a thousand agents, so the orchestrator's only job is to understand intent and route it to the agent that can complete it. She calls it an AI chief of staff. How much autonomy the orchestrator gets depends, as with people, on how long it has worked with a team.

Where to start with automation

Asked where a company should begin, Talat points to the most repetitive workflows: the ones everybody knows how to do, that have become normal, and that are annoying but need to get done. Those tasks give the highest return on investment, she says, and are easy to map because they are done consistently.

A strong candidate is repetitive and does not need human judgment, which she describes as gut feelings, creativity and thinking outside the box. Machines do not have common sense, she says, and make weird decisions when left to judge. Sending an email, checking that a contract has all the information needed, generating an invoice, replying to customer conversations or escalating a work order are machine jobs.

The fastest return she has seen is in sales. Customer service matters as part of revenue generation, because it keeps customers happy and lets a company propose new products. But when a business implements a lead-nurturing and lead-qualification agent, she says, it is immediately getting more sales.

Why real estate fits, and what gets in the way

Automation works best, she says, for businesses built on relationships with a high volume of customer interaction, which is why BRIICK works well with real estate businesses, which hold huge customer databases. In Australia and New Zealand, she says, people sell and buy homes on a seven-year cycle, so knowing that a client bought seven years ago is important information. Busy real estate agents do not always have time to check the whole database and personalize outreach.

Onboarding is another fit. A potential buyer has to complete checks before making an offer, and people get busy, so someone has to chase them for documentation. The problem that keeps resurfacing, she says, is knowledge and data scattered everywhere, too many software products, and processes that are not written down, or written down differently from what people actually do. The bigger the company, the bigger the problem.

Size, software and trust in regulated work

Talat argues that size is no longer a competitive advantage and may be a disadvantage. Customers want what they want, when and how they want it, she says, and will leave a big corporation that treats them like a number for a small company offering a more personalized experience. Highly regulated businesses such as banks have a little more time, in her view, but others have to start treating customers better.

Asked by the host whether companies will eventually manage dozens or hundreds of agents, she agrees, and expects software to be built for agents rather than people: websites that AI can read, instead of apps that users have to learn. People will simply talk to an AI agent that knows how to use the software.

On compliance, she says customer data should never be fed into an open source large language model (LLM); it belongs in a database the company controls, just as it does today in its customer relationship management (CRM) system. For financial institutions, the AI follows a script, as human customer service staff do, and escalates to a human when unsure. Before letting an agent talk to customers, a CEO should insist on a lot of testing, a clear knowledge base and example scripts, and should ask about tone of voice and the escalation rules.

What success looks like

Looking ahead, Talat measures success by customers. She says every customer BRIICK has had since day one is still a BRIICK customer and recommends BRIICK to others, and she hopes for maybe 10 times the current number, all advocates. She most wants BRIICK known for helping real estate businesses focus on building relationships, and describes the company as an orchestrator layer that also helps clients build their AI workforce.

Asked what she would do with unlimited resources, she says she would speed up the product build. BRIICK itself runs on its own system, and while she has a vision of how the orchestrator should be, the team has to build it in sprints.

Her advice for a CEO starting tomorrow morning is to look at the data: where it is, how it is connected, and how to bring it into a central place where all knowledge lives, then clean it. For leaders who feel a competitor is moving faster with AI, her answer is short: get started.

Watch the full conversation

Watch the full conversation above. Sara Valentina Talat says the BRIICK website has more about the company, a contact form and a tab where visitors can build their own agent.

Watch on YouTube →

AI InterConnect hosts conversations with founders, investors and operators building with AI. Explore more episodes on the AI InterConnect podcast page.

Back to Blog

AI InterConnect

AI InterConnect connects founders, CEOs, investors and technology leaders through conversations, market intelligence and growth infrastructure.

Explore our podcast, research and GTM systems to identify opportunities, build authority and accelerate business growth.